Repository navigation
Description
FACTORY FX / SIX FIRST-PARTY VST3 (2026-10-30): Factory FX ADR · Factory FX PRD · manifest schema · merged architecture PR #187 · epic #185. Six original bundled VST3 targets: ArSonEQ, ArSonComp, ArSonVerb, ArSonHarmony, ArSonColor, ArSonKuPik Enhancer. NO binary FactoryFX modules are currently shipped by this docs change; first qualify ASKPVST public v0.5.26 + ArSonEQ feature branch artifacts and obtain product/AGPL/JUCE/EULA approval; ArSonVerb license and Comp/Harmony/Color DSP are later gates. ASKPVST #60, EQ #6, Verb #4. G5A #178 and MS1 #184 are already MERGED. Factory immutable VST3 modules must use existing native FX graph while USER vault remains separate; protect L2 #176, PDC #95, actual hardware #15. Do NOT mark a plugin loadable from a roadmap card.
R6 CURRENT EXECUTION / L2 PRESERVATION — 2026-10-30: Code-audited architecture and priorities · Operator GUI contract · Merged PR #181 · UX implementation #182 · L2 immutable GUI #176. Source baseline main@8251bd6 (re-fetch HEAD before editing): G3 and L2 merged; channelFormat MS1 #173, typed bus MS2 #174, independent hardware outputs MS3 #175 and uniform FX/PDC #95 are NOT completed. #15 real zero-FX audio hardware blocker remains unresolved. Protect ONE seven-row grid and fader alignment for CH1–10+MASTER, FX/ROUTE/SENDS left rail, top MIXER/FX RACK/PATCH, CH1 VST+Orbit, native epoch/revision ACK/MUTE. No old PR bulk-merge, fake Master FX/monitor pins or GUI-only LIVE.
MONO/STEREO REFORM — CANONICAL EXECUTION (2026-10-30)
ADR R5 native MONO/STEREO | R5 handoff | merged docs PR #172.
User product requirement: processing channel MONO/STEREO must be independently editable from physical input source format; bus processing formats and hardware outputs independently typed. Existing ChannelRoutingContract exposes channel format and mono fold, but NativeDryRouteCommand couples format and source and forces BUS stereo. Current G3 send editor merged, not format-aware bus processing. Do not create GUI-only stereo buttons or a second graph.
Order / ownership: MS1 #173 native channel format + GUI → MS2 #174 format-aware sends/FX AUX/GROUP → MS3 #175 physical output patch (historical #143 is CLOSED) → MS4 #95 FX/PDC/Show expansion. #15 real hardware crackle is a parallel release blocker. One vertical native+GUI+test slice per PR, latest main rebase and single integrator. No code change declared done just because docs/issues exist.
Canonical product outcome
Build ArSonMix as a professional, lightweight live digital mixer + portable VST3 host, guided by operator workflows in Waves eMotion LV1 Classic/v17 and Studio One rather than copied artwork. Source: nine screenshots supplied by project owner and the 2026-10-30 YouTube comments export. Reference and auditable criteria are in GUI workflow benchmark and Voice of Customer requirements, PR #169 (docs-only, merged as 7900cb8).
Source of truth / current implementation
- Verified main@fd7a18e (2026-10-30): 10 native inputs + 4 independent stereo internal FX AUX returns, max 32 combined PRE/POST send edges, CH1 8 VST3 insert slots with Orbit, CH2–10 and AUX one routed native insert; verified DIRECT input pin mapping and epoch/revision ACK, MAIN L/R physical output read-only.
- NOT YET arbitrary bus-to-physical output patch, independent MON/CUE, GROUP/DCA/MATRIX, per-CH multi-VST, PDC, fully portable native VST stateful Shows, configurable bounded channel count, multitrack player, cross-platform release.
- P2: restore CH1 VST ADD after Audio Setup, unify BTN-180 plugin UI and eliminate startup white frame #166 CH1 VST ADD merged. G1/G2: FX/ROUTE/SENDS mixer bank and native hardware Audio I/O console #167 G1/G2 MERGED with FX/ROUTE/SENDS and custom INPUTS/OUTPUTS/DEVICE operator UI; full hardware mode/rate/buffer setup, physical multi-output, G3 send controls and real hardware FAT remain unfinished. P0 — Crackling with 0 VST3: diagnose WASAPI clock/input/output buffering before engine expansion #15 0-VST device crackle remains a release blocker. No mock UI or CI can certify physical audio.
Prioritized product gates with measurable acceptance
- P0: audio safety gate (P0 — Crackling with 0 VST3: diagnose WASAPI clock/input/output buffering before engine expansion #15) — actual ASIO/WASAPI no-FX physical continuity, no unmeasured xrun=0 claims, test at representative buffers/clock paths.
- G1/G2: verify merged MIX UX and truthful custom AUDIO I/O (G1/G2: FX/ROUTE/SENDS mixer bank and native hardware Audio I/O console #167) — fixed strips/faders across FX/ROUTE/SENDS; CH1 multi-insert always intact; logical M/L/R matrix; device-epoch and native ACK; no fake output pins.
- G3: usable shared FX Send/Return (ARD P0/P2.6: True per-strip hardware input selection (mono/stereo), channel sends & routable bus graph #109) — at least MIC1 and MIC2 to one AUX1 VST reverb, independently editable send PRE/POST/pan/level, one FX instance, correct return level/fader/mute, real native signal trace.
- G4: real multi-output patch (P0 PRODUCT ROUTING: replace LAB with unified native live mixer + eMotion-LV1-inspired INPUT/INTERNAL/OUTPUT patch matrix #143 extension) — on 8-output device MAIN to 1/2, MON to 3/4 and further qualified bus to 5/6 with independent signal; on 2-output reject unavailable pins; no double-sum/cycle; control-thread prepared plan, no realtime alloc/locks.
- G5: bus/FX depth (ARD: All-channel native FX Rack + safe drag reorder + yellow-green toolbar + recessed fader #95/ARD P0/P2.6: True per-strip hardware input selection (mono/stereo), channel sends & routable bus graph #109 follow-up) — distinguish FX AUX, GROUP audio sum, DCA gain-control-only, MON/CUE, MATRIX and physical out; unified native multi-insert CH2–10/AUX and parallel latency compensation (PDC), no stale plugin lifecycle/Show rollback.
- Later: bounded dynamic channel/bus capacity and configurable banked MIX/ALL view; tested 8/16/32 configuration targets only where benchmark supports them, no "unlimited" advertising.
- Later: playback/performer — audio file player, playlist/setlist, multitrack stems as explicit time-aligned input source nodes, not unclocked browser-only audio.
- Later: release ecosystem — owned private portable VST vault + validated Show recovery, helpful vocal/music/film presets, device/CPU/RAM diagnostics, software compatibility/platform feasibility. Don't promise VST2, Win7, macOS/Android or standalone EXE-as-plugin without independent qualification.
GUI non-negotiables
Fixed meter/fader/mute geometry; operator-friendly compact steel/BTN-180 aesthetic; readable but not oversized fonts/cards; no blur/glow/hover repaint or large DOM reallocations on telemetry. Separate MIX, CHANNEL, FX RACK, ROUTING, SHOW, SETUP; optional ALL and SENDS ON FADERS only with explicit safe mode. Selected source/destination route visualization instead of per-frame spaghetti lines. Raw JUCE dialog advanced only after custom UI can accurately query/mutate native device settings.
Engineering acceptance
A feature is DELIVERED only when native signal behavior, UI state, negative-ACK rollback, exact-head tests, and real-hardware FAT when applicable agree. Stable plugin instances on navigation and FX reorder. Capabilities derive from real devices, not assumed interface names. No second graph/callback, no unmute on patch apply; PDC for wet/dry sum; show save/recall transactional. Test UI at 1366x768, 1440 and 1680 with integrated GPU.
Documentation / merge provenance
PR #169 is merged into main@7900cb889ebec8581daef9d010b95cc7b981a190. Normative GUI and VoC requirements are linked directly from README, AGENTS, PRD v2.0 and ARD v2.0; no third-party screenshot assets were committed.
Ownership / dedup
Use existing #15, #109, #143, #95 and active #167; this is the umbrella PRODUCT TRACKER rather than duplicate bug reports. Refer to PRD v2.0, ARD v2.0, AGENTS.md and PR #169 before planning implementation. New work must reconcile actual latest main and work from complete vertical slices.
Activity
docs(product): verified LV1 + Fender Studio Pro 8.2 parity roadmap an…
masarray commentedon Oct 30, 2026
2026-10-30 benchmark reconciliation — LV1 + Studio One/Fender Studio Pro 8.2 (roadmap committed)
Canonical milestone roadmap: docs/ROADMAP_LV1_STUDIO_PRO_MIXER_PARITY_20261030.md, published to main in docs-only commits
d8d70a6candf820a340. README and AGENTS.md now link it and require verification in subsequent threads. It complements—does not replace—the R6 native/GUI architecture audit #181 and immutable L2 geometry baseline #176.External vendor verified official references: Waves LV1 Classic user guide, Waves Live latency/PDC, Fender Studio Pro 8 Console manual, Fender Studio Pro 8.2 announcement 2026-10-20. Studio One Pro was renamed Fender Studio Pro in 2026; aim mixer/operator workflow parity, not a full DAW timeline or fake SoundGrid/64/80 channels.
Current status:
Dependency order: MS0 #15 parallel (but release blocking) + MS1 #173 → MS2 #174 → MS3 #175; MS4 #95 / G5A #178 foundational graph allowed in parallel with non-overlap/no duplicate native graph, user FX+PDC after exact-head gates; MS5 UI work can start on native states but no invented bus; MS6 data model early and full Show after MS1–4. One real end-to-end native + GUI + negative-ACK + CI + real hardware FAT milestone at a time. No stale PR bulk merges.
This is ROADMAP / SOURCE AUDIT DONE, not feature implementation or hardware qualification. Success requires actual user journeys and exact-head Windows artifact on each slice.
docs(reconcile): lock merged MS1+G5A native delivery and update next …
masarray commentedon Oct 30, 2026
Native delivery reconciliation — MS1 + G5A both merged (2026-10-30)
MS1 PR #184 merged
961c286dafter Chromium 40984090092 and Windows native CTest/verified package 40984090111 SUCCESS. G5A #178 was already merged to main; no direct overlap. This materially delivers independent native processing channel MONO/STEREO separate from physical input, one typed route ACK, native PCM regression, PATCH/MIX GUI, and V4 DRY Show capture/recall migration.Cross-thread documentation permanently reconciled in
AGENTS.md,README.md,docs/ROADMAP_LV1_STUDIO_PRO_MIXER_PARITY_20261030.mdatmain@aa9b9c81. NEXT: MS2 #174 true typed FX AUX + GROUP and format-aware sends, MS3 #175 real physical output patch; G5B #95 real user multi-insert controls + PDC, MS5 #179 operator view and MS6 #180 full vendor Show remain to build. Actual hardware 0VST crackle FAT #15 still OPEN; PR CTest does not prove ASIO continuity. No stale PR bulk merges or fake controls.masarray commentedon Oct 30, 2026
2026-10-30 — FX / Bussing / Engine / GUI parity audit (current main@928dfdfd)
Evidence: AGENTS.md, canonical PRD/ARD/roadmap, source files
PreparedMixerBlock.h,LiveDryRoutingLab.h,LiveChannelFxGraph.h,AuxBusPlan.h,ChannelSendPlan.h; merged native G5A #178, MS1 #184, G5B #186; Windows native main 41012133356 + Chromium 41012133300 SUCCESS.Engine current: one prepared 10 input / 4 stereo internal FX AUX graph; up to 32 PRE/POST sends, deterministic DAG/solo, physical input epoch/revision/MUTED APPLY/native ACK; independent source vs MONO/STEREO channel DSP (MS1). 8 real ordered native VST3 slots in routed CH2–10 and AUX (G5B), alongside original CH1 multi-insert/Orbit, without additional callbacks; real plugin instantiation on control, accepted route lives in DSP.
GUI current: locked L2 11-channel same-grid fader/meters, left FX/ROUTE/SENDS rail, top MIXER/FX RACK/PATCH; G3 native SENDS editor, compact VST plugin picker with favorites/search, actual/acknowledged FX chain; ensure unsupported actions don't masquerade as LIVE.
NOT yet Studio One/Fender Studio Pro or Waves eMotion LV1 functional parity: no full MONO/STEREO FX AUX/GROUP/MONITOR/MATRIX audio bus taxonomy, DCA layer, independently routable hardware MON/CUE/multiout, native post-sum Master FX rack, general VST bus layout negotiation/PDC, vendor state/Recall Safe entire Show, complete six original bundled VST3, or release-qualified low-end hardware CPU/GUI. G5B bypass/editor per routed slot remains limited. Current MAIN physical output L/R only.
Engineering underway (separate non-overlapping PRs):
Next dependency order: (parallel P0 #15 dry zero-FX hardware crackle FAT) MS2 #174 typed GROUP/AUX + correct PRE/POST + VST formats/PDC; MS3 #175 real verified physical output patch; MS4 #95 Master+all-channel FX/PDC and factory binary verification; MS5 #179 selected Channel Focus and sends-on-faders; MS6 #180/#85 comprehensive Show and portable state; MS7 hardware soak/release. Preserve L2 #176 and G5B / single RT graph; no bulk import of stale PR #115/#117/#128 or heavy CI for docs-only updates. Code pass ≠ hardware pass.
masarray commentedon Oct 30, 2026
Cross-thread reconciliation — 2026-10-30 (updated after FXF1 merge)
Current main:
79473ef3c6a94267e3c4b3d60587c6d5f6a192d0. GUI PR #191 MERGED (LCD-integrated dBFS ticks for mono/stereo/MASTER; blue gain-tracking lower fader slit; byte-synced runtime CSS + standalone; no new animation, blur, RAF or native audio changes). Chromium exact-head CI #41033084500 SUCCESS and post-merge main Chromium SUCCESS. L2 #176 common 11-strip/fader geometry preserved. Brushed-metal texture deliberately not activated pending standard/iGPU profiling (#56).Factory FX PR #188 MERGED as
79473ef3: all-or-nothing Factory VST3 manifest registry, SHA-256 file checks, JUCE host UID verification, background discovery/catalog and REAL Factory/My Plugins source filters. Exact-head PR Windows + Chromium SUCCESS; post-merge main Chromium SUCCESS, post-merge Windows still pending at update. No six real FX binaries bundled yet; runtime integrity != publisher authenticity. Release signing, artifact provenance, licensing, original FX build/packaging and audible hardware qualification remain #185/#180/#15.Engine in flight, NOT on main: MS2A #189 exact branch head
9da3a66d, commit fixes all observed legacy integration PCM baselines to new mono PRE sqrt(0.5) pan formula; Windows native CI #41034121201 still BUILDING at update, no green claim. G5C #190 remains draft with exact-head Windows/UI green but needs release/owner gates. DYN0 #193 is independent runtime roster work and must be reconciled before other channel roster edits. Previously merged G5A #178, MS1 #184 and G5B #186 must not be recreated.Delivery priority: hardware #15 zero-FX ASIO/WASAPI crackle FAT and #56 low-end pointer latency; #174 typed AUX/bus send correctness then #175 bus→independent physical outputs; #95 safe per-insert bypass, latency/PDC and master FX; #179 operator channel/bus focus, sends-on-faders and no hidden insert; #180/#85 full native Show and portable plugin state; #185 six approved real Factory FX binaries. Preserve ACK, epoch/revision safe structural apply, zero-heap audio callback, L2 geometry and test evidence. Do not bulk merge stale PR #128/#117/#115. CI green ≠ real 0FX hardware acceptance ≠ functional LV1/Studio One parity.
masarray commentedon Oct 30, 2026
2026-10-30 owner correction — no permanent 10/12-channel product cap: new DYN #192, implementation DYN0 PR #193. All channels must share one actual DSP/FX/routing/UI model; CH1 has to migrate away from the special legacy callback while preserving Orbit/vendor instance state. Stable 64-bit IDs, runtime-sized session, lazy FX/buffers and UI bank virtualization; physical device port count is NOT logical channel count. DYN0 currently is control-only with tested 128/1024 descriptors, NOT audible dynamic channels; DYN1 DSP and DYN2 native-ACKed ADD CHANNEL/Show remain required. Do not implement UI-only + buttons or another audio callback. Protect parallel MS2 #189, G5C #190, FactoryFX #188 and release FAT #15.
masarray commentedon Oct 30, 2026
2026-10-30 delivery reconciliation: exact native merged commits verified: #190 G5C VST3 per-instance BYPASS
593427d(Windows GUI hardware-FAT package in run 41044034992 SUCCESS), #193 DYN0 stable-ID runtime-sized channel roster8e53f1d(exact-head Windows run 41044867738 SUCCESS, stereo add regression fixed), #189 MS2A format-aware mono PRE native sends2c2bdf6(exact-head Windows run 41046000413 SUCCESS). Other concurrent main merges #195 event-driven rack reconciliation and #194 G5D vendor editor have landed; review their exact combined main CI before claiming a fresh cross-feature artifact. Dynamic ADD CHANNEL audio is NOT DONE:stripCount=10currently still constrains one native audio graph; DYN1 refactor must prove true PCM and the unified CH1..N processing path before any user+ CHANNELUI or Show V5 claim. DYN2 banked/virtualized operator UI and stable-ID Show migration remain open. #15 actual ASIO/WASAPI crackle FAT open. No Studio One/LV1 parity claim.masarray commentedon Oct 30, 2026
2026-10-30 — P1 console responsiveness and UI session milestone MERGED (exact evidence)
adc3389d67926923c47dbb829fecd37439a7e4f1.focusChannel; 11 strips/FX anchor/meter nodes do not rebuild merely because focus changes. Full render remains only for actual topology/identity/session changes.main@adc3389dWindows CI #41052042768 and Chromium main CI #41052043000 were in progress when checked, so do NOT label the integrated binary qualified until those complete.docs/P1_GUI_CONSOLE_FOCUS_IDLE_PERSISTENCE_20261030.md.masarray commentedon Oct 30, 2026
DYN1A source milestone delivered as open PR #199 (2026-10-30): runtime-sized C++20 immutable PCM DAG, five real patched input lanes CH1/CH11/CH32/CH64/CH127 feeding MAIN out of 128 configured channel IDs, lazy execution/buffers (no CPU/FX cost for disconnected lanes), 1024-logical-channel scratch budget test, same FX delegate on every active channel, PRE/POST+mono equal-power, MUTE/SOLO, duplicate-send/cycle/invalid-source rollback. Exact head
ddf1567Windows CI run 41051705731 is IN PROGRESS, NOT YET MERGED. Do not call CH11 LIVE in JUCE: oldstripCount=10engine remains authoritative. DYN1B single live graph replacement + CH1 VST3/Orbit handoff and DYN2 native-ACKed Add Channel/Show/banked GUI remain required. MS3A #198 and G5E #197 concurrent; no shared mutations in #199.masarray commentedon Oct 30, 2026
2026-10-30 — Integrated native operator FAT RC / current main reconciliation
Code baseline merged in order: MS3A #198
f2f192dphysical sparse MAIN L/R output APPLY+ACK, DYN1A #199d31d44eruntime PCM core offline only, G5E #1977124e22routed CH/AUX VST3 binary state Show Save/Restore; MS3A2 #20120cd7878same leased physical MAIN pair feeds FX de-click and LUFS/dBTP, eliminating the HW1/2 hardcoding when output is routed elsewhere.Exact testing and actual qualification: MS3A2 exact-head Windows Native/CTest 41057398228 SUCCESS; integrated main Chromium 41059050351 SUCCESS; integrated main Windows 41059050360 in-progress at this comment, NOT certified until successful Windows build, full CTest and SHA-verified ZIP upload. Code milestone source SHA
20cd7878614e90135efeb23d45784c7fb832b1aa; documentation-only PR #202 merged as0801cc8with zero need for additional native CI.Operator test pack: current Windows FAT guide includes actual ASIO/WASAPI 0 FX tone/capture/dry, sparse 2/2 & 8-output port routing, accepted vs rejected revisions, post-selection fade/LUFS, routed CH2/AUX1 VST3 state parameter A→B→A, negative hash/catalog tests, standard PC interaction and structured evidence. Prominent override added to historical packaged FAT guide and README. Native Show remains incompatible with the separately borrowed CH1 legacy FX rack; no expectation of arbitrary plug-in vendor state portability across PCs.
Next CODE milestone should be DYN1B (not more GUI mock rows): integrate
DynamicPreparedMixerBlockinto the one JUCE callback replacing fixed 10-input graph, with exact FX instance ownership, CH1 Orbit legacy migration, all 4 AUX, typed mono/stereo sends, physical output pair, native epoch/revisioned MUTE/APPLY/ACK, Show V4 safe migration and RT zero heap/lock. Only then DYN2 adds genuine native-ACKed channel ADD/REMOVE/DUPLICATE + virtualization/banks + Show V5 stable ChannelIds. Gate 10→11→32→128 audible PCM, FX distinct instances, PRE/POST, solo/mute, sparse inputs, stale device, memory budget failure rollback and vendor state. Do not claim 128 LIVE channels today. Separate #15 0FX crackle hardware FAT and low-end GPU p95 remain release gates.masarray commentedon Oct 30, 2026
M5 premium console GUI DONE — native CLIP, recessed blue slit, VST legibility
Merged PR #204 to
main@5d7761357c702a4aaadebb2ac8e4a2ad4c965e94. Exact final PR head50f29b2735818f722d838732b89b7386cd796d85passed full Chromium run #41063147292, including L2 geometry and native VST3 mocked UI.--fader-open. No blur/filter/rAF/hover handler.#91EFB6, leaves muted/bypassed rows gray and all separately clickable power buttons intact. Preview use 17px child inset; real native VST3 button already has 16px inset so its label adds 6px only, preventing double indentation.docs/M5_FADER_CLIP_VST_CONSOLE_FAT_20261030.md.main@5d776135Chromium #41063384039 and Windows Native/CTest/package #41063384068 were in progress at the latest check. Do not offer the new native ZIP as PASSED until exact-main Windows run succeeds. Standard iGPU pointer p95 and ASIO 0FX crackle P0 — Crackling with 0 VST3: diagnose WASAPI clock/input/output buffering before engine expansion #15 remain hardware FAT release gates. MS3B MS3B: true simultaneous MAIN + MONITOR physical buses with 14 POST sends and single native graph #203 is independent native output DSP work; do not confuse it with M5.masarray commentedon Oct 30, 2026
M5 integrated CI correction — PR #207 merged, avoid invalid Windows artifact
After M5 #204 was merged, Chromium exact-head PR #41063147292 and first main #41063384039 passed, but main Windows #41063384068 FAILED in Node source contracts before MSVC. Four historical tests expected literal
percent(d.inputPeak),percent(d.peakLeft),percent(d.peakRight); M5 had refactored them to equivalent prevalidated local amplitudesraw/rawL/rawR. NO audio/GUI semantic or C++ failure was involved. #207 restores direct native telemetry source expressions while retainingrawPeak>=1native CLIP latch args, and updates one matching M5 assertion. Exact PR Chromium #41063879155 SUCCESS; merged tomain@ddad5c73257fb34ca72773637f9cfcc346c2856e. Main integrated CI: Chromium #41064172092 and Windows Native/CTest/package #41064172504 were in-progress when checked. Download ONLY latest successful exact-main provenance-matching artifact. No duplicate DSP engine, no extra RAF/filter, no false hardware-certification claims.masarray commentedon Oct 30, 2026
Integrated L2 operator fix merged via PR #208 #208 as main@9746e3ee71562ff08e017cef9613d66593342aef. Exact-head Chromium 41070701536 SUCCESS. ROUTE was 163px upper (148px compact), overpacked input/output/format/patch; now 224px normal and 195px compact, with all CH1–10 and MASTER faders shrinking uniformly. CH2–10 native VST3 empty insert plus glyph and ADD FX label, plus separate FX power chrome, now match CH1. MASTER displays six equally high native FX rack slots, all deliberately disabled (true Master VST3 engine not integrated; no false live FX). Canonical standalone CSS synchronized. Playwright checks all 11 strips and ROUTE bounds in 4 viewport sizes, six disabled Master slots, native FX bypass. Postmerge main Chromium https://github.com/masarray/arsonmix/actions/runs/41071015423 and Windows Native/CTest/package https://github.com/masarray/arsonmix/actions/runs/41071015432 running at this comment; exact-main artifact NOT qualified until success. Physical #15 0FX crackle and standard-PC mouse p95 unresolved. Next engine milestone needs same-graph dynamic native VST3 Master ownership, not visual-only controls. Engineering guide: docs/L2_ROUTE_DENSITY_NATIVE_FX_PARITY_20261030.md
masarray commentedon Oct 30, 2026
REKONSILIASI EKSEKUSI AKTUAL — 2026-10-31 (source/CI verified)
Single current baseline:
main@9746e3ee71562ff08e017cef9613d66593342aef, includes #204 GUI fader/CLIP, #207 native meter regression, #206 real Studio-like M/L/R hardware I/O matrix, and #208 unified 11-strip ROUTE/FX rack/L2 layout. Exact-main tests Windows Native/CTest/package #41071015432 SUCCESS; Chromium #41071015423 SUCCESS. GitHub Pages publish #41071143435: first deployment failed upstream HTTP 500 (artifact build succeeded); reran only the failed deploy job, attempt 2 SUCCESS, no costly MSVC rerun. CI GREEN != actual ASIO/WASAPI hardware/FAT PASS.Reconciled milestone truth (MAIN, not historical snapshots)
ONE ACTIVE NEW-ENGINE INTEGRATION OWNER: MS3B PR #203, exact PR head
8a574aacb5e31c920024cf4e9196544152360818: Windows native/CTest #41069188480 and Chromium #41069188256 both SUCCESS. But this PR was prepared againstad6e4cafand is now 1 main commit behind #208 (code/standalone and smoke/geometry overlap). It is DRAFT, NOT merged.mergeable=truedoes NOT replace source-level integration review. Next action: reconcile #208's full 11-strip GUI, standalone parity, geometry/meter contracts and tests into #203 without losing physical MAIN+MONITOR matrix, then one exact-HEAD Windows/CTest and Chromium gate; merge only after both GREEN. No premature claim that MONITOR is present on main.IMMEDIATELY AFTER MS3B: DYN1B one meaningful vertical slice replacing the fixed graph inside the same live callback by the stable-ID dynamic graph; must preserve 14 CH/AUX FX lanes and vendor/Orbit Show identities, PRE/POST send semantics, sparse MAIN/MONITOR physical destinations, MUTE→APPLY→epoch/revision ACK, realtime no-heap/lock, fader ramps, meters, and rollback. Prove CH1 parity and live CH11+ real PCM with vendor FX before DYN2 +CHANNEL + virtualized bank/Show V5. Do not build a second audio engine or cosmetic +CHANNEL.
PARALLEL, NON-OVERLAPPING: #15 hardware 0FX crackle/capture/output diagnostics; #174 GROUP/PDC planning; #185 sibling original plugin releases. Don't independently edit MS3B's
LiveDryRoutingLab.h,LiveAudioEngine.h,PreparedMixerBlock.h,LiveChannelFxGraph.h,web/mixer-io.js/css,web/monitor.js, standalone HTML, or IO smoke tests while its integration owner is active.PR HYGIENE: 19 open PRs at audit: #203 is the current MS3B integration candidate, the other 18 are historical/overlapping (#1, #3–#10, #25/#26/#29, #59/#60, #113/#115/#117/#128). QUARANTINE them from merge; source-audit each before any close/reuse. Never bulk-merge, force replace latest
main, delete owners' work or rerun Windows for docs-only changes. Historical AGENTS/PRD/ARD/roadmap paragraphs referring to stale SHAs or "MS1 not delivered" are snapshots, NOT today's acceptance evidence. This timestamped source+CI audit supersedes those status claims only; preserve their product requirements.Release acceptance labels kept separate: MERGED CODE ≠ exact SHA CI GREEN ≠ operator hardware FAT = ACCEPTED. No user auto-unmute, fake pins, unsupported Master FX or unqualified zero-latency claim.
2 remaining items
masarray commentedon Oct 30, 2026
2026-10-31 current-main and PR convergence: PR #215 DYN1B-2A merged at
ff173964; PR #217 canonical convergence ledger merged at71bcd715. The approved RAIL3/L2 GUI, native MAIN+MONITOR, OUTPUT1 and FactoryFX release admission are preserved. 18 pre-R6 open PR branches are divergent and quarantined, not bulk-merged or deleted; ledger maps each to its current product owner. Next priorities: #15 physical 0-FX crackle FAT (parallel P0), #192 DYN1B-2B one-callback cutover, #95 all-channel VST3/PDC, #179/#182 true LV1 operator workflow, #180 portable Show/Recall Safe, MS7 low-end GPU/device/plugin soak. Passing GitHub CI does not certify real hardware or zero roundtrip latency.masarray commentedon Oct 31, 2026
2026-10-31 CROSS-THREAD PRODUCT RESET — canonical decision now merged: docs/PRODUCT_TRUTH_CROSS_THREAD_20261031.md via PR #221, commit
91115f37.Owner-facing diagnosis: merged technical milestones have not produced one coherent live soundman workflow. PATCH1 device-safe but STAGE/MUTE/APPLY-heavy input routing was superseded by PATCH2 docs only (#220); MS5 Sends-on-Faders (#219) currently edits route drafts, not direct live send levels; Channel Focus/group/DCA and native MASTER FX parity are incomplete; DYN1B remains offline; 0FX crackle issue #15 physical qualification remains open.
mainhad newest native MSVC/CTest/ZIP success498ded667run #41152162311 but this is not hardware FAT.ONE priority contract (M0/M1/M2): M0 real 0FX wired ASIO/WASAPI FAT and iGPU; M1 native live operator SOURCE→FX→SEND→BUS→OUTPUT, realtime safe gain/send/solo/pan distinct from structural MUTE/APPLY (PATCH2 + MS5), chosen channel and AUX master focus; M2 one live dynamic ChannelId/VST/Show graph, all channels parity/virtualized banking, FactoryFX legal releases. Keep output user-selected, never OBS/virtual-only; protect one callback/controller, epoch/revision ACK, L2/RAIL3 and OUTPUT1/MS3B. Do not count docs, mock-driver UI, offline dynamic PCM, green Windows CI or cosmetic patches as finished mixer journeys. Mandatory new-thread entry point is now at top of AGENTS and README.
masarray commentedon Oct 31, 2026
M1 NATIVE LIVE AUX GAIN MERGED (31 Oct 2026) — PR #223 squash commit
e11a9799d5926567bedb8e4c8549ddc1d7b6dd62. PR exact-head Chromium #41155087583 and Windows native MSVC/CTest/package #41155087488 SUCCESS. New LIVE parameter pathroutingAuxSendGain/arMixAuxSendGainAckadjusts ONLY existing CH→AUX POST send coefficient during audio playback; 32 bounded atomics, callback-only <=64-sample smoothing, no global MUTE, plugin recreation, route revision increment, audio device restart or MAIN fader mutation. Browser reports LIVE level only after native ACK. Structural CREATE/REMOVE, PRE/PAN, AUX enable/driver pin still use guarded MUTE/APPLY. M1 operator/FAT contract. Next real-device candidate from post-mergemainWindows #41155784392, currently running; do not claim its binary exists until uploaded verified artifact. Physical #15 remains OPEN. This is the FIRST M1 realtime send-gain vertical slice, not a declaration that PATCH2 selected-channel quick-source, AUX master fader flip or full LV1 parity is complete.masarray commentedon Oct 31, 2026
M1 VERIFIED WINDOWS HARDWARE-FAT PACKAGE AVAILABLE (31 Oct 2026):
main@e11a9799d5926567bedb8e4c8549ddc1d7b6dd62, post-merge Windows MSVC/CTest/packaging #41155784392 SUCCESS and Chromium #41155784393 SUCCESS. Artifacts →ArSonMix-GUI10CH-P012-Verified-Hardware-FAT-Windows-x64. Downloaded full ZIP, CRC PASS,BUILD_PROVENANCE.jsonexact sourceCommit matchese11a9799, EXE SHA2561a5fadce9dd3e4768d3c2d43926a5b585e013aa1c8d0dd0069a3fadb7601e9eaindependently verified. Now contains new operator LIVE AUX POST send gain in native EXE, not just preview or old M0 baseline. FAT: WIRED same-interface DIRECT ASIO/WASAPI, 0FX output-only/capture/dry first; create POST→AUX while outputs MUTED, APPLY once, then unmute and change SEND FADERS level/keyboard without further global MUTE or APPLY. Monitor for click/pop and verify MAIN fader untouched. Physical #15 remains unresolved until audible/diagnostic results; the package is EXPERIMENTAL FAT, not stage-certified. See M1 contract.masarray commentedon Oct 31, 2026
2026-10-31 live PR reconciliation: canonical engineering ledger is PR #229 /
docs/PR_BACKLOG_RECONCILIATION_20261031.md(branch until merged). Audited main443760319324, merged #223 live POST AUX sends, #224 SOURCE1+64-pin actual JUCE device, #227 custom device UI docs-only; #222 closed superseded. Historical 18 open PRs are quarantined, not current sprint tasks. Active #226 selected CHANNEL + AUX MASTER and #228 custom DEVICE1 share Main.cpp/mixer-io/standalone/tests: do not merge independently on stale bases, revalidate same-head Windows+Chromium and preserve single native audio graph/ack. FactoryFX old #225 was conflict-blocked and is being selectively replaced by #229 without audio owner changes. #15 physical 0FX crackle remains a release blocker. GitHub mergeable/CI green ≠ physical FAT or zero-latency certification.masarray commentedon Oct 31, 2026
M1 integrated operator milestone merged 31 October 2026: PR #231 → main
d3293af72eef029032a9138f26e6c71dc69967f7after reconcile of #228 custom ASIO/WASAPI device setup and now-closed superseded #226. Exact integrated PR head12582f29passed Windows MSVC/CTest/ZIP #41163412113 and Chromium UI #41163412112, not just separate-branch tests.Genuine operator journey: custom hardware driver/source/output ports → selected CHANNEL accepted SOURCE→FX→POST SEND→OUTPUT, single native controller/ACK, live CH→AUX POST send gain (#223) and independent AUX MASTER gain on selected bus. Four atomic bounded post-AUX-FX targets smoothly affect BOTH MAIN and MONITOR; no duplicated DSP callbacks, no runtime plugin re-instantiation or global mute for level changes. AUDIO I/O custom device work preserved in same files with exact standalone HTML module sync. #226 closed as superseded to prevent stale remerge.
Release gates still open: post-merge main Windows run #41164371465 in progress at initial check, only provide its verified ZIP after SUCCESS; hardware P0 #15 crackle 0FX unresolved; native MASTER insert/PDC, dynamic live channel >10, group/DCA/matrix, real device FAT and full professional parity remain unavailable. Do not call this user-certified full mixer.
masarray commentedon Oct 31, 2026
2026-10-31 M1 SOURCE OPERATOR JOURNEY merged: PR #233 ->
main@a7fb61511ffe27fdef1f9923695daf6899646dff, exact PR-head Chromium UI smoke run #41166533918 PASS (20 steps); same WebView2/standalone source CSS ownership and operator regression. Selected CH SOURCE now opens compact chooser from native-enumerated enabled pins, accepts mono/adjacent stereo, sends one existing guarded native route transaction and waits for ACK; nonadjacent stereo enters ADVANCED PATCH. This deliberately refuses armed-output source edits, because click-free source crossfade is not yet verified. No second JS audio graph, new virtual pin, or fake live ACK. Combined with #231 native device setup + CHANNEL SOURCE/FX/SEND/OUT focus + live AUX master, and #223 continuous POST send parameter, the basic M1 operator flow is materially closer but still NOT OPERATOR-READY: native source handover while armed, bussing/GROUP, master FX/PDC and hardware zeroFX issue #15 remain open. Post-merge Windows run #41166662476 and Chromium run #41166662407 in progress; do not release-qualify without exact-head CI and physical FAT.masarray commentedon Oct 31, 2026
2026-10-31 verified cross-thread handoff — SOURCE2/DEVICE2
main@a7fb61511ffe27fdef1f9923695daf6899646dff: M1 SOURCE2: selected-channel verified input chooser using one native route ACK #233 merged, selected-channel SOURCE opens actual driver-verified port chooser and commits using existing mutedroutingDryApply/ACK. Do not call this unmuted clickless source switching or named logical alias registry.0d75fce9, but GitHub rejected merging after M1 SOURCE2: selected-channel verified input chooser using one native route ACK #233 due conflicts. Rather than force/overwrite, new M1 integration: DEVICE2 guarded hardware scan + merged SOURCE2 operator flow #234 replays the five source/test changes on latest main as152ae0bd6e0c760af52d3a1afc13ba82290c6b41. Scope: cached native ASIO/WASAPI inventory, explicit guarded rescan, actual opened masks/rate/buffer readback, WebView2/standalone parity. M1 integration: DEVICE2 guarded hardware scan + merged SOURCE2 operator flow #234 Chromium CI PASS; Windows native CI is pending at time of this comment. Do not merge the stale DEVICE2: cached hardware inventory, guarded driver rescan and verified active I/O masks #232 directly. Review M1 integration: DEVICE2 guarded hardware scan + merged SOURCE2 operator flow #234 exact-head CI before integrating and only then close/supersede DEVICE2: cached hardware inventory, guarded driver rescan and verified active I/O masks #232.web/monitor.jsalone is the route authority. Next genuinely major work: operator-safe live source handoff and unified per-channel VST/FX lifecycle, while physical P0 — Crackling with 0 VST3: diagnose WASAPI clock/input/output buffering before engine expansion #15 FAT runs as a release gate. Avoid splitting these into another duplicate UI engine.masarray commentedon Oct 31, 2026
M1 integration verified and merged (2026-10-31). PR #236 on exact head
259331438d2dc0666c9449bb527602ae7e0cbfb8passed Chromium 41175414889 and native Windows MSVC/CTest 41175414858, merged to main@0835806b89ecef650faa989a2545c52a0b2885aa. This completes a meaningful M1 software integration: true in-drag native ACK/backpressured CH→AUX POST send faders and AUX master, selected-channel named logical MONO/STEREO sources tied to verified backend+enabled pin mask, and same-source-format live physical input switch via the one native graph with a pre-FX 64-sample PCM crossfade + device epoch/route revision native ACK. Initial unpatched input, format change, hardware/FX/structural graph edits still require safe MUTE/APPLY/ACK. Main Windows 41176318876 is producing a merged SHA artifact; exact user hardware FAT, #15 crackle/soak, latency and full product acceptance remain NOT QUALIFIED. Fixed live 10CH+4AUX and other M2 work still open. Attention parallel M2 #235: its older head touches web/monitor.js and tests/mixer_io_smoke.py; rebase/replay only unique M2 changes onto main@0835806b, preserve M1 SOURCE/AUX/DEVICE and do not bulk merge stale controller. No duplicate DSP or GUI route authority.masarray commentedon Oct 31, 2026
2026-10-31 M1 operator completion checkpoint (source-verified).
main@708795888e04172a080e6a738bdabc12d5f58b7fnow contains #234 DEVICE2 (safe cached hardware inventory), #236 M1 (named hardware input/source same-format native 64-sample switch, coalesced live AUX send/master ACK), #235 M2 BUS/FX (existing BUS group/parallel FX routing), and #238 M1 focused SEND editor live-level alignment. #238 was merged after exact-head Chromium UI smoke SUCCESS (run 41196791338) with native-only gain API for accepted POST sends and enabled AUX returns, guarded structural PRE/ON/OFF/PAN/hardware APPLY retained; native and standalone HTML JS parity. The last pre-#238main@54e4229Windows/CTest/ZIP SUCCESS (run 41195859441); latestmain@70879588Windows/Chromium post-merge checks are in progress at this note, not yet a qualified artifact. Physical ASIO/WASAPI, device disconnect/soak, and #15 zero-VST crackle remain UNQUALIFIED. 10 input+4 AUX native graph remains actual LIVE scope; do not call full eMotion LV1 product parity or hardware acceptance complete. M1 workflow has a substantive native code path, but user acceptance requires wired interface FAT evidence. Follow current main, not dated PRD snapshots.masarray commentedon Oct 31, 2026
Canonical Studio One Console audit and M3–M5 product gate (31 Oct 2026): see source+current Chromium screenshot audit and single-graph architecture decision in #192. Audited exact
main@91043355b876(Windows & Chromium green). DO NOT mistake M2B's borrowed unused 10CH BUS or identical FX CSS for Studio One parity. User expects native runtime Add Channel/Add Bus/Add Bus for Selected Channels (true independent ChannelId, no physical pin consumed), exact one strip/FX engine, one-click multi-select subgroup, primary OUT vs parallel SEND clearly visible on EVERY strip, BUS input-source count+popovers and selected route highlighting, bounded prepared RT graph/one callback, actual Show portability. Maintain premium compact ArSonMix aesthetic, 1366px banking, low-end responsiveness; no bulk cards/glow/ghost browser-only audio. Product priorities: M3 dynamic native graph/FX legacy migration, M4 Console workflow/route visibility, M5 PDC/MASTER FX/physical #15 FAT. Do not start another cosmetic GUI PR before real engine source of truth.masarray commentedon Oct 31, 2026
2026-10-31 LIVE HARDWARE FAILURE — DO NOT CLAIM OPERATOR-READY. User screenshot shows Windows Audio capture
CABLE Output (VB-Audio Virtual Cable)→ renderSpeakers (Realtek), 48 kHz/480 frames, CH1/CH2 patch + native ACK yet no audible PCM. Verified native 10CH callback hadinputPeak_ = 0hardcoded;SIGNAL CHECKINPUT painted routed MAIN peak, anddiagnosis()returned 'route ready' before examining mute, callback or raw input. Output remains deliberately muted after device/routing APPLY, so a green route label cannot establish audibility. VB-CABLE CABLE Output is recording side, needing a source player directed to Windows playback CABLE Input. PR #241 introduces raw capture PCM peak before native master MUTE, provenance and device-epoch sample blocks, compact CAPTURE/ROUTE/OUTPUT status in Audio I/O + existing Signal Check button, and explicit source guidance with independent native status. Chromium CI success on6b7197b9; Windows MSVC/RT CI still building at posting. Do not call #15 physically closed, do not auto-unmute or reroute drivers silently. The next real acceptance is playback program → CABLE Input → ArSonMix CABLE Output → Realtek, with capture PCM nonzero while safe-muted, native route ACK, low-volume ARM and audible clean 0-FX output. Follow exact head of PR #241 and issue #15 physical steps, not prior build screenshots.masarray commentedon Oct 31, 2026
P0 audibility integration merged 2026-10-31: PR #241 is merged as
main@19ebb4663590168e12d5d60d7d6700b3da3b47bf. Exact PR-head Windows native + realtime/packaging CI SUCCESS (run 41203805443); Chromium SUCCESS (run 41203805411). Main post-merge Windows native CI run 41204771528 is currently compiling to produce accessible FAT ZIP. Changes: real raw capture input peak prior to native MUTE early exit, independent CAPTURE→ROUTE→OUTPUT diagnostic UI, proper Signal Check input bar and VB-CABLE routing explanation. This is software verified, not physical audio success; issue #15 remains pending user's on-device input/arm/output/zero-FX tests. Preservemonitor_fail-closed safety. Do not replace it with automatic unmute or browser DSP.masarray commentedon Oct 31, 2026
ISSUE RECONCILIATION (2026-11-01). Closed as duplicate/superseded (NOT a 'feature done' assertion). October product-epic priorities and old Studio One/LV1 release matrix are superseded by canonical PRD v5 and central live engine issue #192, which now owns full end-to-end Console UX and native graph. Retain this issue as historical product discussion archive.
Single product/engine authority: PRD v5, ARD v5, #192. Physical audible quality: #15. This ticket's prior evidence remains searchable.